Weird buzzing sound after updating to 2022.24.6 [Fixed in 2022.36.1]

[WilliamG]

to be honest other manufacturers has either A) no updates at all b) if they did update, they will never allow you to downgrade due to the fact their framework is more of a firmware baked into the chip that is PITA to write. I had issues with bothe Cadillac garbage CUE software and Audi MMI updates.

but I agree competition will make everyone better, so hopefully better quality builds and better Service Center experience.

Sure, but we expect better from Tesla, - as we should, and they're massively dropping the ball on what is their highest-end vehicle (well, X grabbed that crown recently, but still...). These are their largest margin vehicles, so you'd think it might behoove them to have a single software engineer and hardware manager monitoring these messageboards, and reporting back to the mothership. But apparently not.

 

[kazamm]

Sure, but we expect better from Tesla, - as we should, and they're massively dropping the ball on what is their highest-end vehicle (well, X grabbed that crown recently, but still...). These are their largest margin vehicles, so you'd think it might behoove them to have a single software engineer and hardware manager monitoring these messageboards, and reporting back to the mothership. But apparently not.

This is exactly it

 

[NeuroRad]

to be honest other manufacturers has either A) no updates at all b) if they did update, they will never allow you to downgrade due to the fact their framework is more of a firmware baked into the chip that is PITA to write. I had issues with bothe Cadillac garbage CUE software and Audi MMI updates.

but I agree competition will make everyone better, so hopefully better quality builds and better Service Center experience.

Ford ota updates on my 21 f150 were flawless.

 

[uscbucsfan]

Ford ota updates on my 21 f150 were flawless.

Ford only pushed SYNC 4 updates OTA until recently. They are toying with more integral updates currently, but Mach E have received very OTA updates (maps from what I was told), they still have to go to service for major updates involving the battery, driving, etc. My neighbor has one.

It's not a good comparison.

 

[NeuroRad]

Ford only pushed SYNC 4 updates OTA until recently. They are toying with more integral updates currently, but Mach E have received very OTA updates (maps from what I was told), they still have to go to service for major updates involving the battery, driving, etc. My neighbor has one.

It's not a good comparison.

I received bluecruise OTA which worked fantastically.

 

[Ash83]

People are reporting they needed to go the dealer: BlueCruise OTA update

Anyways not the right place to discuss other OEMs here.

 

[uscbucsfan]

People are reporting they needed to go the dealer: BlueCruise OTA update

Anyways not the right place to discuss other OEMs here.

Apparently it's VIN based and limited to who can get them OTA, but you are correct...it's off topic. Even if Tesla is the only one doing complete OTA updates, not doing enough quality control/product verification is an issue. This isn't the first update with bugs.

Almost every major update this year has had a bug....USB issue, Dashcam/Summons at the same time, this, and many more that I can't remember of the top of my head.

 

[vackachan]

Never had this issue but today when me car got 2022.24.6 i noticed after the update this werid buzzing sound coming from the car


I did reset and power off/on with no luck

Opened service request with Tesla and waiting there reaponse but wanted to see if anyone experienced this.

Because the sound seems (not 100% sure coming from the frunk toward the lower side) i am worried something wrong with the battery so i moved the car outside the garage away from my other Tesla waiting for the service center to respond.

I hear the same sound that you have recorded while parked in my garage or in quiet places from my Long Range S that I got on August 10, 2022. I reported this problem to them on September 10 and scheduled the service for September 21. Since they may not hear the sound in their noisy shop, I asked them for a mobile service so that they could hear the sound while parked in my garage. I also told them that I could send the recorded voice of this sound to them. Tesla responded by saying that they could not do a mobile service for this problem and I have to bring the car to them for service. I had multiple communications with them from September 10 to 21 regarding this issue. But they never informed me that this was a known problem. However, when I took the car to them on September 21 at the scheduled time, the tech said this is a know problem for S and X and he would fix the problem in 15 minutes by updating the "firmware" and asked me to wait for it. He came back in about half an hour and told me that he could not fix it and it will be fixed in the next release. Then I asked him why Tesla asked me to come to their shop for this if they already knew about this problem and wasted my time and their time.

 

[Pauliwog]

I have already contacted the SC manager and they are aware of the issue. Nothing they can do other than feed my feedback on downgrading to the developers.

In reference to downgrading that is no more or less difficult than a new release. There is no need to regression test an old release already which was operational prior to a new FW update. You're not introducing new code which has not been already operational on the cars various ECMs.

Without access to the source I can't definitively say how the updates are performed or stored but it would not be difficult to store the previous FW in a sperate partition of NVM for downgrade. Needless to say this half assed approach of pumping out untested or poorly tested code against HITL is just par for the course. The lack of SW quality assurance is right inline with their HW QA putting the service centers in a bind fixing post production issues.

I am just trying to bring the issue to their attention. Spend $150K on a car that does this 24/7

 

[af88]

I have both a new S plaid and X. If it is not this it is other issues they cause.

This latest release was a "bug fix" release, just like all the other prior releases they introduced new bugs.

It is beyond frustrating when you spend this kind of money to have to play a game of cat and mouse with HW or SW failures, then resort to internet search to find you are in good company. Service centers have no clue either what the issues are rather than telling Freemont to get their act together.

How there has not been a class action lawsuit against Tesla up to this point is beyond me on these new refresh models.

 

[Ash83]

I am just trying to bring the issue to their attention. Spend $150K on a car that does this 24/7

Yeah, what most us been doing last few weeks since we encountered this is to disable Sentry mode when parked home so the car can go to sleep ( keep key fobs away and do not open the App) and when the car goes to sleep the aound will stop.

Hopefully 2022.36 will fix this mess.

 

[af88]

True. I don't see why they couldn't push an old software. But, they'd have to figure out which of the many other MCUs would also need to be downgraded to be compatible. And the existing old software release package might not have those firmwares in it, since they weren't changed as a part of the upgrade. I.e. Some door controller MCU might have been upgraded by a newer firmware, but it needs to be downgraded now, but the old door firmware isn't present in the last release since it was last updated 10 releases ago, or some other nonsense.

But, agreed. Quality control, particularly around NVH is awful at Tesla. And same with manufacturing quality. That's why I've been back to the service center six times to fix issues noted at delivery. It's coming up on a year and I still don't have all the issues fixed on my Plaid. This is my fifth Tesla and the worst manufacturing experience so far, so I'm pretty sure this is my last.

That is a valid point on non updated MCU, but nothing they can't resolve by simply looking at the CSCI or SVD to determine what MCU was updated in the release and download the prior version OTA. If capacity is an issue get rid of these absurd games no one plays, although im sure most of the update are stored in flash.

Point is there is a way to offer this, and the argument I got was invalid prior making this request. These issues exacerbate the need to roll back.

 

[zoomer0056]

the argument I got

What was the argument you got?

 

[af88]

Grasping at straws they came up with security concerns. Reality is at any given time there is half a dozen builds available with only a small percentage of cars with the most updated release.

If there is a real vulnerability, it won't be an optional update, nor would they allow owners to skip updates.

The fact is they had no good excuse, it is always easy to throw the security flag up. I was merely looking for them to roll back the FW for a test to validate the issue was FW related, which it was. They are well aware most of these release are bug riddled, even the service center doesn't question it anymore because they won't listen to their feedback.

 

[Maybe22]

Definitely my last Tesla and I tell everyone who asks me.

 

[whitex]

In reference to downgrading that is no more or less difficult than a new release. There is no need to regression test an old release already which was operational prior to a new FW update. You're not introducing new code which has not been already operational on the cars various ECMs.

To be fair, that is not true. Most manufacturers do not allow downgrades for security reasons. Imagine you had a software release which can be hacked and allow the attacker to control your car (perhaps drive it into the wall of your garage remotely). Tesla patches it, sending out a new release, but allowing downgrades, all the hacker has to do is to deploy a downgrade and off into your living room we go. Hyperbole, perhaps, but it illustrates the point why security typically mandates disabling downgrades. There are other technical reasons too, downgrades also need to be tested to make sure there are no weird interaction, for example your MCU downgraded first but some other firmware in the system is of higher version, now the MCU can no longer talk to this "future" firmware, not even to downgrade it. Also, while upgraded, perhaps some files got written using newer formats, when you downgrade, they look like corrupt files to the old firmware.

Bottom line, downgrades are expensive to support and potentially insecure, hence most manufacturers do not support them. Tesla would have build a new release here with the old source code, then thoroughly test to make sure this "upgrade" applies correctly and does not cause any issues (apparently their process to do so missed this buzzing issue in the first place, so it's not full-proof). It probably is easier and quicker to root cause the buzz and only revert that code, than release a patch, but they may choose to just wait till next release as it may be scheduled soon - famous Tesla service answer "this should be fixed next update, if not, the one after that".

Note by the way that I am not defending Tesla's continuous integration/test on customers release model. While it works for Tesla, it sucks for customers who have to live with issues like that for weeks or months. I am just stating technical reasons why downgrades are technically more challenging than a new release.

 

[Ash83]

One more charge last night and battery dropped another 0.1%

Was 0% for 6 months before updating to this garbage 2022.24.6 and since then it dropped by 0.5% in less than 3 weeks.

 

[BamaATL]

Utter absurdity we can't roll back SW and have to deal with this after just doing an update and coming on here to find out I'm not the only one. f you Tesla

Good news is that you AREN'T the only one. That would be worse. It's annoying but I was glad to know most everyone was having the same issue. A software update will fix and if everyone is having the issue, it will get fixed sooner.

 

[af88]

To be fair, that is not true. Most manufacturers do not allow downgrades for security reasons. Imagine you had a software release which can be hacked and allow the attacker to control your car (perhaps drive it into the wall of your garage remotely). Tesla patches it, sending out a new release, but allowing downgrades, all the hacker has to do is to deploy a downgrade and off into your living room we go. Hyperbole, perhaps, but it illustrates the point why security typically mandates disabling downgrades. There are other technical reasons too, downgrades also need to be tested to make sure there are no weird interaction, for example your MCU downgraded first but some other firmware in the system is of higher version, now the MCU can no longer talk to this "future" firmware, not even to downgrade it. Also, while upgraded, perhaps some files got written using newer formats, when you downgrade, they look like corrupt files to the old firmware.

Bottom line, downgrades are expensive to support and potentially insecure, hence most manufacturers do not support them. Tesla would have build a new release here with the old source code, then thoroughly test to make sure this "upgrade" applies correctly and does not cause any issues (apparently their process to do so missed this buzzing issue in the first place, so it's not full-proof). It probably is easier and quicker to root cause the buzz and only revert that code, than release a patch, but they may choose to just wait till next release as it may be scheduled soon - famous Tesla service answer "this should be fixed next update, if not, the one after that".

Note by the way that I am not defending Tesla's continuous integration/test on customers release model. While it works for Tesla, it sucks for customers who have to live with issues like that for weeks or months. I am just stating technical reasons why downgrades are technically more challenging than a new release.

Your argument is completely invalid for the reasons I already explained in my previous posts. I'm not going to debate common sense and hypotheticals with you. It is pretty clear you don't have a cyber or SW background. Carry on

 

[WilliamG]

Good news is that you AREN'T the only one. That would be worse. It's annoying but I was glad to know most everyone was having the same issue. A software update will fix and if everyone is having the issue, it will get fixed sooner.

Except that’s not always been true (see the vibration thread). But we live in hope.